This research shows how to apply the Reaction Null-Space method for controller design of a humanoid robot such that the balance and motion/force control subtasks are dynamically decoupled. The performance of the controller is verified via simulations, using the model of a miniature humanoid robot HOAP-2.
